Neuronal damage after stroke varies, resulting in differential response to restorative therapies despite similar presentation of motor deficits. Greater insight into neuromechanisms associated with recovery is one strategy that may improve prediction of stroke recovery and maximize therapeutic effects by informing targeted interventions. The long-term goal is to develop a competitively funded research program investigating neural substrates of motor impairment and recovery in stroke, with an emphasis on walking function. The overall objectives in this application are to evaluate (i) sensorimotor network connectivity as a determinant of treatment-related walking improvement, (ii) associations of treatment-related alterations of white matter integrity/connectivity with walking gains, and (iii) relationships of neurotrophic and growth factors with response to neurorehabilitation. The central hypothesis is that structural connectivity of key sensorimotor brain regions beyond the lesion site supports the plasticity necessary for motor recovery after stroke. The rationale for this project is that determination of neural and vascular biomarkers of walking recovery post-stroke will likely offer a strong scientific framework to assess potential for recovery and inform targeted interventions. The proposed objectives will be pursued through the following specific aims: 1) Elucidate the effect of post-stroke preservation of sensorimotor structural connectivity on response to gait training in chronic stroke; 2) Evaluate the relationship between treatment-related changes in white matter integrity and walking gains following neurorehabilitation; and 3) Elucidate relationships of circulating levels of brain-derived neurotrophic factor (BDNF) and vascular endothelial growth factor (VEGF) and response to gait training. A pre-post intervention research design will be utilized for this study. Twenty-eight persons with chronic stroke will be recruited from local clinical facilities and University stroke research databases. Select inclusion criteria include >6 months post-(ischemic) stroke and ability to walk 15 meters (50 feet) with occasional moderate assistance for balance; select exclusion criteria include uncontrolled hypertension and contraindications to MRI. Subjects will receive intensive training 2 hrs/day for 10 days across 2-3 weeks; each session will devote one hour to overground gait training and one hour to balance activities. Clinical measures of strength, balance, and mobility; detailed gait analysis; blood samples for serum BDNF/VEGF levels; and MRI scans will be performed pre-post intervention. Blood samples will also be obtained mid-way through the intervention to examine changes in circulating levels of BDNF/VEGF across the intervention. Using neural and vascular biomarkers to predict response to gait training in chronic stroke is innovative. The proposed research is significant because it is expected to provide strong scientific support for future clinical trials designed to target therapies based on predicted functional potential via MRI and serum markers. Ultimately, such knowledge has the potential of offering additional mobility gains to those likely to benefit from treatment, even in the chronic stages of stroke.